With the development of network information, the network development of an enterprise also progresses constantly. Generally, IDC (Internet datacenters) network of enterprises are ubiquitous across many different regions. Moreover, multi-layered network devices are existent inside an IDC and inter-IDCs; therefore, there is a big challenge to monitor link quality of an entire network. First, the link data of the entire network is very huge, such that it is generally hard to monitor all links; second, because the round-trip paths of an asymmetric network are inconsistent, an end-to-end delay cannot be accurately tested; finally, when the network fails, it is very hard to quickly locate a failed link.
The existing network link monitor technology, although it may confirm that congestion occurs in a certain port of a certain device at the time by collecting device port traffic, it cannot confirm the congestion condition of the entire traffic data flow, and cannot measure an end-to-end delay. A generally employed traceroute tool at two ends of the server has the following problems: it is unable to monitor all links of the whole network; it is unable to solve the issue caused by an asymmetric link; and it is unable to locate a problem quickly.